The development of genetic vectors has heralded the fast-growing field of somatic gene transfer. (Anderson, W. F., Science, 1984, 226:401-409). Vectors based on simple retroviruses, such as the Moloney Leukemia Virus (MoMLV), are often selected because they efficiently integrate into the genome of the target cell. Integration is thought to be a prerequisite for long-term expression of the transduced gene. However, efficient gene transfer to tumor tissue has been a major impediment to treatment of cell proliferative disorders despite the use of viral vectors such as retroviruses.
In the early steps of infection, retroviruses deliver their nucleoprotein core into the cytoplasm of the target cell. Here, reverse transcription of the viral genome takes place while the core matures into a preintegration complex. The complex must reach the nucleus to achieve integration of the viral DNA into the host cell chromosomes. For simple retroviruses (oncoretroviruses), this step requires the dissolution of the nuclear membrane at mitotic prophase, most likely because the bulky size of the preintegration complex prevents its passive diffusion through the nuclear pores because there are no nuclear localization signals to facilitate active transport into the nucleus.
Currently retroviral vectors used for human gene therapy are replication-defective and must be produced in packaging cells, which contain integrated wild type virus genome sequences and thus provide all of the structural elements necessary to assemble viruses (i.e., the gag, pol, and env gene products), but cannot encapsidate their own wild type virus genomes due to a deletion of the packaging signal sequence (psi). Replication-defective virus vectors created by removal of the viral structural genes and replacement with therapeutic genes are introduced into the packaging cells; so long as these vectors contain the psi signal, they can take advantage of the structural proteins provided by the cells and be encapsidated into virion. However, after infection of a target cell, the vectors are incapable of secondary horizontal infections of adjacent cells due to the deletion of the essential viral genes.
The use of replication-defective vectors has been an important safeguard against the uncontrolled spread of virus, as replication-competent retroviruses have been shown to cause malignancies in primates (Donahue et al., J. Exp. Med., 1992, 176:1124-1135). However, replication-defective retroviral vectors are produced from the packaging cells at titers on the order of only 106-7 colony-forming units (cfu) per ml, which is barely adequate for transduction in vivo. In fact, clinical trials for gene therapy of glioblastoma multiforme, a highly malignant brain tumor, have encountered major problems in achieving adequate levels of tumor cell transduction, and despite promising initial results in animal studies (Culver et al., Science, 1992, 256:1550-1552). In order to increase transduction levels as much as possible, instead of using a single shot of virus-containing supernatant, the virus packaging cell line PA317 itself was injected into the brain tumors to constitutively produce retrovirus vectors carrying the HSV-tk gene (Oldfield et al., Human Gene Therapy, 1993, 4:39-69). Subsequently, the protocol was further modified to include a debulking procedure followed by multiple injection sites, as it was found that the virus vectors did not diffuse far enough from the site of initial injection. Despite these modifications, the transduction efficiency has been estimated to less than 1% of the tumor cell mass and any significant tumor destruction is presumed to be due to the potent bystander effect of the HSV-tk/ganciclovir treatment. Thus efficient transduction of cancer cells in a solid tumor mass represents a major problem for cancer gene therapy.
Accordingly, there is a need for a gene transfer vector capable of high-level transduction in vivo, while limiting uncontrolled spread of replication-competent virus which could result in insertional mutagenesis and carcinogenesis.